© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 9367
    Prev
    Next
    INDEX
    Explanations

    The neuron is firing on proper names and other capitalized named‐entity tokens (e.g. “Me” as a divine pronoun, “Apple devices,” “Cavanaugh,” verse-initial “We,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    و
    0.99
    س
    0.95
    ات
    0.93
    фак
    0.91
    та
    0.90
    ار
    0.86
    ア
    0.84
    s
    0.81
    MARGIN
    0.81
    ल
    0.80
    POSITIVE LOGITS
     distilling
    0.89
     lete
    0.88
     PAOK
    0.87
     teve
    0.86
     resveratrol
    0.86
     diatomic
    0.86
     tinge
    0.84
     resonant
    0.83
     shivering
    0.83
     carbonic
    0.82
    Activations Density 0.000%

    No Known Activations